The relationship between Banco Galicia and BeBot dates back many years, but in the last 3 years, we have been working on the design of Gala, the bank's chatbot.
The goal
Initially, when the team consisted of three Product Analysts and two implementers, users needing to address specific queries were often directed to the contact center. The goal was, and continues to be, to put Gala and the rest of the digital channels at the forefront, reserving telephone support for sales and complex issues, and minimizing branch-related inquiries to zero.
We accompany them in a consultancy that includes:
Team development: Together with the design and content managers, the Human Resources team, and the Product Owner of the Contactability area, we professionalized the role of Conversational Designer within the bank and ensure their professional development and organizational growth.
Product strategy: The Product Owner, the Content Manager and us contemplate what Gala will be like in six months, one, three, and five years. We define actionable steps, identify potential obstacles, and begin working to steer her towards that future.
Content improvement: Together, the BeBot and Galicia teams define, design, and implement the content in the bot, ensuring that all content pieces are consistent in personality and writing. We form a team of up to 5 people dedicated 100% to the design of conversation architecture and content.
Team development: Trainning & Role Definition
When we started working on this milestone, the team assigned to the bot design had no formal knowledge of conversational product design, lacked a direct technical leader, and did not belong to any design area. So, our job was:
Training technical profiles: design and deliver training for the internal Product Analyst team and accompany them on their journey to become the first Conversational Designers at the bank. We taught them everything we know about conversation architecture, writing, personality, and product design
The program explained how to design the structure of a conversation using flowcharts and how to write with personality.
Suporting the leaders: Collaborating as external consultants in the creation of the Content Manager position, and coaching their leaders during their first year in charge of the Content COE (Center of Excellence)
Placing the team in the right pot: Assisting the Content Manager and the Human Resources team in defining the role of Conversational Designer and integrating them, first into the Content COE and later into the technology team.
Product strategy: Conversatonal Design process
We designed and implemented the conversational content design process and the continuous improvement process within the Contactability team.
The biggest challenge we faced was incorporating the design process into the existing bot development process. Until then, implementers and developers received summarized content from a specific area of the bank and implemented it into the bot. There was no process to oversee content hierarchy or curation. Personality did not exist, and no one tested the consistency of voice and tone. It took several months of presentations, numerous discussions with the technology, implementation, and analysis teams, and long hours of coaching and support for the Conversational Designers team to start establishing a presence in the bank.
Estrategia de producto: Proceso de mejora continua
Product strategy: continuous improvement
This process came much later because initially, we didn't have enough metrics to measure content in production. The first thing the Business Analytics team did, guided by their Product Owner, was to tag each message that the bot sent to users within a flow and then create funnels. The initial metrics we started measuring were user drop-offs within flows with high traffic: cards and accounts. As the bot grew, we also incorporated new metrics, such as C-Sat and resolution. The Product Analysts team began creating dashboards where we could visualize all this data and start making decisions about the content.
Mejora de contenidos: El Diseño de la personalidad
Content improvement: personality
Designing Gala's personality was a significant challenge because Galicia has many clients, and each one is different. From those who only contact the bank to check if they received their salary to those with a high-risk investment profile. The customer who uses the bank to finance domestic purchases is also different from the one who needs to finance machinery to improve the production of their fields.
To encompass all these users and realities, we designed a personality with three clear and distinct attributes:
Passionate: It focuses on the positive and enthusiastically encourages you to go beyond. It is optimistic and always highlights the positive side of things. It's a voice that encourages you to always move forward.
Professional: It's a voice that exudes confidence in the words it chooses. It is experienced, and it shows when explaining banking topics
Natural: It speaks like the customer and sets aside banking terminology. When we write, we choose words to demonstrate that you can rely on her.
The goal of writing with personality is to be coherent and consistent. Giving personality to a bot is about creating patterns in the conversation that convey confidence to the user. We carry out this work with constant asynchronous feedback from the team and two weekly work sessions, the Content Review and the Design Critique.
Conclution
It was a long journey we traveled with Gala, but it serves as an ideal example to explain the evolution of a conversational product. Designing and implementing a bot in three months to serve a user segment is a good start, but that's just the first step. For the bot to become a product that provides real value, there's much to be done. From designing processes so that all teams can work smoothly to designing the information architecture of each conversational piece. Designing a bot requires the same commitment and investment of money and time as any other technology product.